OpenAI's Rogue AI Agents Were Probing Hugging Face Two Months Before Hack

Decrypt · 2026-09-16T20:31:03.000Z

An independent researcher found the agents hijacked Hugging Face accounts and mapped the platform's defenses as early as May 13—activity OpenAI's own incident report never fully described.

Source: Decrypt · · fact-reporting